油藏描述中的地震相分析新技术.ppt

油藏描述中的地震相分析新技术 油藏描述中的地震相分析新技术 绪论地震相分析技术在油藏描述中的作用波形分类地震相分析技术的特点和应用实例 以Stratimagic软件为例地震相分析技术存在的问题和发展趋势 绪论 地震相是个 古老 宽泛的名词 概念 60年代中期 有人开始使用 70年代初随着地震地层学的兴起 被广泛使用 地震相的定义 多种多样不统一 本人愿意定义为 地震信号特征的一种表征形式 并且这种表征形式所表征的信号特征可以在横向或纵向上划分成单元或分类 地震信号外形 内部结构 振幅 相位 频率 速度等均可以作为地震相 现在一般将振幅 相位等表征地震动力学特征的信号称为属性 而把反射外形等静力学特征的信号定义为地震相的较多 地震相可应用于地震地层学 岩性地震学以及油藏描述中的储层预测等 本次讲座主要集中在地震相的概念 以及工业界已应用的地震相分析方法的原理 实例的介绍上 油藏描述中的地震相分析新技术 绪论地震相分析技术在油藏描述中的作用波形分类地震相分析技术的特点和应用实例 以Stratimagic软件为例地震相分析技术存在的问题和发展趋势 地震相分析技术在油藏描述中的作用 油藏描述的任务 储层 油藏分布预测 储层 油藏物性的确定解决油藏描述的地震方法 属性 地震相分析方法 井约束AI EI反演方法推荐的工作流程 构造解释 沿层选定目的段 层段内地震相分析 靶区的井约束反演 地震相和反演结果的综合解释 定性到定量的地震相分析 WhatisResultsfromtheSeismicFaciesAnalysisTechnology ComparisonwithTraditionalMethods MapofAmplitudeStandardDeviation thebestAmplitudeMap SeismicFaciessuperimposedontheAmplitudemap EACHrealtraceisassignedacoloraccordingtowhichmodeltraceitismostcloselycorrelated TheSeismicFaciesMap 地震方法用于油藏描述的现状 AttributiesAnalysis 地震相地震属性分析方法所提取属性种类不断增加 20 50种 更多 用户选择属性缺少合适的方法对多种属性解释 地质意义不明确 WellCalibrationandInversion地震的井标定和反演外推估算地震信号的横向变化通常是困难的需要先验的初始模型花费和计算吞吐量仍是系统化工业化应用的障碍先验约束往往出现误差 地震相分析技术在油藏描述中的作用 快速进行地震信号特征的分类 研究地震信号的变化规律从地震信号某种 多种特征的变化规律中确定反映地质体沉积 物性等变化的规律 从而直接进行沉积相研究 储层预测和物性预测等 地震相分析可以快速的为井约束AI EI反演等确定靶区 指导反演结果的解释 新的地震相分析方法可以进一步确定地震微相 地震相的定量化等 从而进行油藏的精细描述 Well1 Well2 Example ChannelDefinition 1 ConventionalInstantaneous AverageAmplitudeMaps Example ChannelDefinition 2 SeismicFaciesMapofisolatedchannelregion usingaNeuralNetworkderivedclassification Model 12classes B A Example ChannelDefinition 3 DetailofCentralChannelDifferencesinproductionfromwellsAandBareexplainedWellB central clean channelWellA pointbar 地震相分析技术在油藏描述中的作用 快速进行地震信号特征的分类 研究地震信号的变化规律从地震信号某种 多种特征的变化规律中确定反映地质体沉积 物性等变化的规律 从而直接进行沉积相研究 储层预测和物性预测等 地震相分析可以快速的为井约束AI EI反演等确定靶区 指导反演结果的解释 新的地震相分析方法可以进一步确定地震微相 地震相的定量化等 从而进行油藏的精细描述 油藏描述中的地震相分析新技术 绪论地震相分析技术在油藏描述中的作用波形分类地震相分析技术的特点和应用实例 以Stratimagic软件为例地震相分析技术存在的问题和发展趋势 波形分类地震相分析技术的特点和实例 AttributiesAnalysis 地震相地震属性分析方法所提取属性种类不断增加 20 50种 更多 用户选择属性缺少合适的方法对多种属性解释 地质意义不明确 WellCalibrationandInversion地震的井标定和反演外推估算地震信号的横向变化通常是困难的需要先验的初始模型花费和计算吞吐量仍是系统化工业化应用的障碍先验约束往往出现误差 Stratimagic 地震地层解释 地震相分析软件 专门用于解释岩性 地层 油藏 地质相对比的新的地震解释技术源于ELF公司获得专利的波形分类技术 由CGG FLAGSHIP开发为软件产品 2002年Paradigm购并Flagship后 进一步与其它的地震相分析技术结合 如Seisfacies NexModel VoxelGeo等 使其更加完整 功能强大 Stratimagic auniquesolutionStratimagic 独特的解决方案波形分类地震相分析 Aprocess characterizationbasedontraceshape一种处理 基于道形状的特征描述Traceshapeclassificationrepresentsthetrueheterogenityoftheseismicsignal道形状分类代表了地震信号的真实的横向异常Atechnology self organizingneuralnetworks一项技术 自组织的神经网络Anindustrialshape recognitionprocess robustandunaffectedbynoiseorspuriousevents一个工业化的形状识别处理 它稳定 不受噪音和假同相轴的影响Amethod manyyearsofoperationalsuccessapplied 一种方法 成功地应多年toexploration appraisalandreservoirstudiesinclasticsandcarbonatesforoilandgas onshoreoroffshoreon5continents fromsea bottomto20 000ft 可用于勘探评价和油藏研究 碎屑岩和碳酸岩 油或气 陆上和海上 TheBasicAssumptionis Changesinanyofthephysicalparametersofthesubsurfacearealwaysreflectedinachangeinshapeoftheseismictrace Forexample changeinporositywillresultinadifferentlyshapedtrace shape isquantifiedinthechangeofsamplevaluefromsampletosample WhatistheSeismicFaciesClassificationTechnologyMentionedHere Whatdoyousee Yourbrainisaneuralnetwork SHAPEisusedtodecidehowmanydifferenttypesofvegetablearehere NOTcolor howmanypeppers orsize howmanytomatoes Arethesethesameshape NowWhatDoYouSee WhatistheSeismicFaciesClassificationTechnologyMentionedHere HowtounderstandthemeaningofseismicdatathroughFaciesIdentificationandClassificationusingTraceshape XX amplitude 2ms Samplingtonearest4mssamplegenerates 2msunbiasednoiseontimeupto25 biasednoiseonamplitude FIXEDVERTICALSAMPLING ReducessamplingnoiseTakesfulladvantageofpropagationbeyondseismicsample TRACERECONSTRUCTION TraceReconstruction acriticalstep WHATAREBASICNEURALNETWORKS SignalFlow InputOutput Synapse INPUTSEISMICINTERVAL OUTPUTTRACES Dendrites CellBody Synapses Axon Lookingforseismicshapechanges NeuralNetwork Clusteringanalysis ANeuralNetworklooksforasuiteoftracesthatdescribetheprogressivechangesintheseismicshape Lookingforseismicshapechanges NeuralNetworkorderedcolorchanges Clusteringanalysisabruptcolorchanges WhatDoWeClassify Wholecube Significantlyexceedsactualvolumeofinterest reservoir goodforearlyexploratoryworkonlyAttributemaps Demandspriorknowledge canbeusedtorefineinsight butnottodefineitProblem Whichmapstouseasinput Problem SomeinformationcouldbebypassedTraceshapeininterval FocusedongeologicalvolumeofinterestSeismicsignalshapeincludesallattributes ComparisonofBenefitsandDrawbacks WhatistheSeismicFaciesClassificationTechnologyMentionedHere HowtounderstandthemeaningofseismicdatathroughFaciesIdentificationandClassificationusingTraceshape 工作流程 workflow I Learningfromthedata andonlythedata从地震数据中学习 且仅仅从地震数据 Themodeltraces模型道Thesesynthetictracesareconstructedbytheneuralnetworkprocess usingalearningsetextractedfromtheseismicinterval Nowelldataisusedatthisstage Theuserhasnoinfluenceontheselectionofdata andtherearenoweightingcriteria Theresultis100 repeatable 这些合成道是用从地震层段中提取出来的由神经网络处理建造的 这一阶段不需要井数据 用户在数据选择方面没有影响 没有加权标准 结果 100 可重复 INPUTSEISMICINTERVAL OUTPUTTRACES Synapses Dendrites CellBody Axon WhatareBasicNeuralNetworks SignalFlow InputOutputSynapse TheProcess TheNeuralNetworktrainsitselfontheactualtraceshapeswithina3Dseismicinterval andconstructssyntheticseismictracesthatrepresentthesignaldiversityovertheentiredefinedvolume Tracesarerefinedbyaniterativeprocessuntilthebestcorrelationtotherealdataisobtained TheSeismicFaciesMap EACHrealtraceisassignedacoloraccordingtowhichmodeltraceitmostcloselycorrelatesto NEURALNETWORKPARAMETERS Numberofmodeltraces numberofcoloursintheoutputfaciesmap NumberofiterationsRateoflearning epsilon Continuity sigma Referencesurfaces intervalthickness sub samplingparameter OUTPUT INPUT PROCESSING ClassificationMaps ClassRange2to100 3Classes 7Classes 15Classes Increasingthenumberofclassesresultsingreaterdetail Smallnumberofclassesidentifiesfirstordertracevariability Unlikeclustering NeuralNetworksdonotrequirepreconceivedideasaboutthenumberofclasses NumberofIterations Range1to100 1Iteration 20Iterations 50Iterations 100Iterations CLASSIFICATIONMAPS Theseismicfaciesmap地震相图 Themap地震相图Eachtracehasbeenassignedthenumber andcolor ofthemodeltracetowhichithasthebestcorrelation 每一道赋给它与模型道最相关的号码和颜色 Byobservingthedistributionofcoloronthismap wecanassessthedistributionofseismicshapesthroughouttheinterpretedarea 通过观察图上颜色的分布 我们可以评定解释区域的地震形状的分布 反映了岩性 地层 地质相的变化 Projectingfaciesinformationonseismic将相的信息投影到地震剖面上 Theclassificationresultcanbeprojecteddirectlyabovetheintervalonwhichtheprocesswasapplied allowingaone to onevisualizationoftheactualdatatracesandtheircorrespondingassignementtooneoftheclasses 分类结果可以直接投影到处理过的层段的上 允许一对一的实际数据道及其中一个相应的赋值分类的可视化 为地震相的变化确定其具体反射特征 利用专门的解释工具 ReflectorTermination Envelops 等 逐线解释出地震相变化的位置和形状 如上超 下超 不整合等 利用Termination和Envelope解释 利用Termination和Envelope进行解释 Wheredowegofromhere FittingthefaciesmaptowellinformationTherelevanceofthefaciesmap s relativetoageologicalsettingcanbeassessedbyfittinginwellinformation IntervalscopeTheprocessisobviouslysensitivetodatathatisincluded orexcluded fromthevolumeofinterest Whileintervalsshouldbelargerthanthestricttime thicknessofinterest tocatche g tuningeffects itisinterestingtotrydifferentthicknesses AreaofinterestOnceageneral purposemaphasindicatedsomemajorfeatures theprocesscanbefocusedonthezonesofinterest toobtainasharper moredetailedpicture II 将地震相结果与井信息匹配进一步细分地震相Fittingwellinformation 与井信息匹配 Seismicsignalatwellposition 井位置的地震信号Foreachwell selecteitherasyntheticseismogramfromalist ortheactualseismictraceatthepenetrationoftheinterval Correlationtoallmodelsiscomputed andacolorisassignedaccordingtobestcorrelation 对于每一口井 选择或者合成记录 或者层段位置的地震道 与所有的模型道计算相关 按照最好的相关赋给颜色 FittingWellInformation ComparingSeismicResponse Realtraceatwelllocationiscomparedwiththemodels Whichisthebestmodel Whereelsecanweseethismodeltype Substitutingtracesinthemodeltable 替换模型道 Stratimagicallowstheusertosubstituteoneofthemodeltraceswiththetracecurrentlyintheselectorwindow beitaseismictracefromthedataset orasyntheticseismogramcomputedfromthewelldata Stratimagic允许用户用现有选择窗口的道替换一个模型道 可以是来自数据体的地震道 或者是由井数据计算的合成记录 Recalculatingonanareaofinterest重新计算关心的区域 Torevealmoredetailinthechannel wecouldincreasethenumberofclasses However itismoreefficienttoprocessanewintervalrestrictedtotheprospectivearea 为了揭示河道内更细的细节 可以增加分类的数目 而更有效的方法是在限定的区域内处理新的层段 III QuantifingSeismicFacieswithPetrophysicalparameters地震相的定量化 岩石物性参数模拟 NexModel PilotedseismicfaciesanalysisusingNexModelTMandStratimagicTM NexModelTMBasicWorkflow Loadlogdata Createlayeredimpedancemodelforwell Loadseismicatwelllocation Createwellsynthetic Tiewellsynthetictoseismic optimise Basic Advancedmodelling Calibrate quantifyseismicfacies Exportnewfaciesgroups SUPERVISEDCLASSIFICATION UNSUPERVISEDCLASSIFICATION reducingrisk addingvaluetosubsurfaceevaluation Apowerfulstartingpoint TheNexModelsyntheticiscorrelatedto ThenearbyseismictracesTheStratimagicNNTtracemodels DT RHOB GR SyntheticSeismogram Seismictracesatborehole withhorizons Synthetictrace StratimagicbestfitmodelTrace Acousticimpedencecolumn NexmodelNexmodel Stratimagic ModeledtraceA ModeledtraceB GuidedclassificationinStratimagic throughmodeltracessubstitutionwithNexmodelsynthetictraces Insertionof thewellAsynthetictrace thepseudo wellsynthetictraces NexmodelNexmodel Stratimagic Nexmodel Stratimagic WellA FaciesA WellB ModeledtraceA ModeledtraceB Othermainfunctions ConventionalAttributesanalysis常规的属性分析工具3Dpropagation3D自动追踪VoxelGeo3D可视化StratiQC成图ConventionalInterpretationtools常规的解释工具 Add onValueandProjectdatabaseintegration高附加值与其它数据库的高度集成 Preservingyourinvestment保护你的投资YourcurrentOpenWorksorGeoFrameprojectdatabasesareaccessed你现有的数据库是OpenWorks还是GeoFrame Stratimagic都可以访问 包括地震数据体 层位和井数据 Youruserscontinuetousetheirfamiliarstructuralinterpretationtools你的用户继续使用他们熟悉的构造解释工具Puttoworkimmediately立即投入工作Withoutduplication accesslarge volumeseismicdatasetsoverthenetwork 不需要备份 可以通过网络访问大的地震数据体 沿层或层间提取多种属性 作为岩性 地层 油藏解释的辅助手段 沿层或层间提取14大类30多种属性 得到属性平面图 作为描述岩性变化等辅助特性 利用叠合图 MixMap 手段 将所有属性结果 地震相图等任意叠合在一起 利用方便的色彩管理工具 突出它们的共同特征 定性的综合对比 为解释岩性 地层 油藏等提供更多的依据 Dipmap finalchannelbase Zoomedarea Channelthicknessattributes Thisattributehighlightsthefeaturesofinterestthathavebeenseenonothermaps Averageamplitudeofpeakeventsininterval Faciesmap channelareaalone Runningthefaciesclassificationprocessontherestrictedareaofthechannelsyieldsamorerepresentativecharacterization Seismicfaciesmap12classes MixMap Faciesmap Avg Ampl Peaks Geologicalmodel prospectivity Prospectivefeaturesweremappedasenvelopes 3Ddelineationofvolume ChannelleveePaleo channelPointbar Enlargementonattributemapofaverageamplitudeofintervalpeaks Randomprofile TimeSlice1660ms Stratimagic2 0NewFeatures Multi AttributeSeismicFaciesClassificationusingNNT Multi volumetraceshapeclassificationonconstantandnon constantintervalsMulti attributemapclassification IntegrationwithVoxelGeo ViewingandmanipulationofStratimagicdatainVoxelGeowithnoworkflowdisruptionsSeismicvolumesSeismicfaciesandattributemapsWells boreholes tops logs SeisFacies Multi attributeseismicanalysis IntroducinganewsoftwaresolutionjointlydevelopedbyENIAgipDivisionandFlagshipGeo 多属性全数据体分类 SeisFaciesBenefits Faciesdistribution anditsassociatedgeologicalmeaning canbeappliedfor PreliminaryscreeningofseismicdatainexplorationactivityRankingofprospectsConditioningofgeostatisticalreservoirmodellingReservoircharacterizationGeohazardriskassessment CLASSIFIEDSEISMICTRACESORSAMPLESCANBEDIRECTLYCALIBRATEDFORQUANTITATIVEDEFINITIONOFRESERVOIRPROPERTIES SeisFacies SeisFaciesincorporatestechnologiesandmethodsdevelopedbyENIAGIPDIVISION Classification Calibration Fusion 3DSEISMICTRACECLASSIFICATIONMULTIATTRIBUTEMAPSCLASSIFICATIONMULTIATTRIBUTEBLOCKCLASSIFICATION SeisFaciesClassificationProcesses ClassificationNNT Hierarchical BLOCKS MAPS TRACES InputData Multiple3DseismicvolumesVariableorconstanttimeinterval InputData IntervalattributemapsHorizonmapsClassificationmaps InputData Multiple3DseismicvolumesVariableorconstanttimeinterval PCA Recommended OutputData 3DSeismicFaciesVolume ClassificationHierarchical OutputData AttributeFaciesMap ClassificationHierarchical PCA Optional OutputData SeismicFaciesMap Zonation Optional PCA Optional Amplitude PCA1 PCA2 Coherency Impedance PCA2 PCA1 PCA2 PCAComponents Amplitude Coherency Impedance InputSeismicVolumes PCA1 SeisFaciesPCA APPLICATIONOFREGRESSIONFUNCTION S CONTROLVARIABLE DEFINITIONOFREGIONS OUTPUT CALIBRATEDVOLUME SeisFaciesMulti AttributeCalibration Classification SEMBLANCEVOLUME SEMBLANCEMAP IMPEDANCEMAP IMPEDANCEVOLUME MIIXATTRRESULTS MIXEDVOLUME MIXEDMAP RESERVOIRCHARACTERIZATION TheSeisFacies Fusion Approach TurbiditeSystem BaseofTurbidite HorizonSlice FaciesBlock HorizonSlice FusionSemblance Impedance Semblance Impedance SeisFaciesFusionExample SeisFaciesConclusions IntegratedcomponentandextensiontoStratimagic sharesStratimagic suserinterfaceandinfrastructureRobustsolutionformulti attributeclassificationandcalibrationofseismicdata incorporatingtechnologiesandmethodsdevelopedbyENIAGIPEnableseffortlessworkonmultipleversionsofaseismicsurvey orasetofattributescomputedovertimeEnablesadetaileddescriptionofthereservoir resultinginbetterinformedbusinessdecisionsbasedonmoreaccuratepredictionofreservesImprovesreservoircharacterizationwithinfielddevelopmentprojects ANewMethodologyBasedonSeismicFaciesAnalysisandLitho SeismicModelingTheElkhornSloughFieldPilotProjectSolanoCounty California ByManuelPoupon FlagshipGeosciences todayParadigm andKostiaAzbel CGG Geoscience Offshore March1999 TheElkhornSloughFieldPilotProjectSolanoCounty California ScopeoftheProjectTheData WintersPinchout3DThePlay DeepWaterFan ChannelStructuralInterpretationConventionalHorizonAttributesGeologicalHorizonAttributesStratigraphicInterpretationConventionalIntervalAttributeAnalysisSeismicFaciesAnalysisModelingSeismicFaciesTraceRevisedGeologicalModelandBusinessImpactsConclusions Interpretationofhorizons faults Before After SeismicClassification TheMissingLink Calibrationtowells InversionorIntervalmapanalysis Geostatisticalanalysis modeling Interpretationofhorizons faults Analyzesurfaceattributes SeismicFaciesClassification Intervalattributeanalysis Interpretationofgeologicalshapes Geostatisticalanalysismodeling TheData WintersPinchout3D 3DSurvey SolanoCo CaliforniaShotandprocessedbyCGG Americasin1995 830in lines 700cross lines 110 x110 binspacing 52sq miles Sampleinterval2msec Recordlength6sec WellData4wellsdrilledonaturbiditicplay basedon2Dinterpretation WellC 70FeetofnetpayintheAsand Recoverablereservesincreasedto14 18BCF ThePlay DeepWaterFan Channel WintersSands TheWinterspinchoutplayisturbiditicinnature sandsbeingtransportedthroughchannelsincisedintotheshelfanddepositedintodeepwaterfanssurroundedbyshales K Lanning1998 RefiningtheChannel FanArea Wecandotwothings 1 Limittheareaofanalysistothechannel fanonly2 Usethewellboretracetopilottheseismicfaciesmap WellA 15ftgasWellB 45ftwaterWellC 70ftgasWellD nosand A B C D Timemap TopofWinterssand Mixedmap Time Dip Azimuthmap StructuralInterpretation ConventionalHorizonAttributesSedimentlayersdiptowardtheSouthWest DipandAzimuthmapsrespectivelyhighlightthechannelandfansystem StratigraphicInterpretation ConventionalHorizonAttributesHorizonAmplitudeMap Twodifferentgeologicalenvironmentsareexpressedwithsimilarhighamplitudevalues Brightspots ThisNeuralNetworkTechnologyislicensedfromTotalFinaElf SeismicFaciesAnalysisusingNNT WhatIsIt SeismicFacies Thedescriptionandgeologicinterpretationofseismicreflectionpatternsincludingconfigurations continuous sigmoidal etc frequency amplitude andcontinuity NeuralNetworkTechnology NNT Theabilitytoanalyzeandclassifytraceshapesusingadiscriminatingprocess SeismicFaciesMap Thisisasimilaritymapofactualtracestoasetofmodeltracesthatrepresentsthediversityofvarioustraceshapespresentinaninterval ModelTraces Intervalofinterest StratigraphicInterpretation UnpilotedRegionalSeismicFaciesAnalysisClassifyingthe60 msintervalabovethereferencehorizonusingNeuralNetworkshaperecognition SeismicfaciesmapshowsturbiditesdepositedalongaNNW SSEpaleo coastline Severalchannelsincisingtheshelfcanalsobeidentified StratigraphicInterpretation UnpilotedChannelSeismicFaciesAnalysisClassifyingthe60 msecintervalbelowthereferencehorizonusingNeuralNetworkshaperecognition Seismicfaciesmaphighlightstheoutlineofanasymmetricfan SeismicresponseatWellC MainstreamNWtoSEortiltedseabottomtowardsSE Asymmetricfan ModelTraces 15 25 70 0 SeismicFaciesMap StratigraphicInterpretation PilotedSeismicFaciesAnalysisUsingtheseismicresponseatWellCasanindicatorofgas chargedsandsandfocusingoverthechannel fanareaonly pilotedseismicfaciesmaphighlightsthedistributionofthethickerreservoirsands SeismictraceatWellCusedasmodel 9 Modeltraces PilotedSeismicFaciesMap Onlyalimitedareainthechannel fansystemhaveseismicresponsessimilartoWellC thicksand WellA B thinsand areoutofthemainfanarea WellD shaledout hasadistinctseismicfacies Petro AcousticModeling ModelingSeismicFaciesTraceUsinglogtracesfromWellCandmodeltrace 9 seismicresponseiscalibratedas70 ofgas chargedsands Petro AcousticModeling SeismicFaciesCalibrationUsinglogtracesatWellCandDtorespectivelycalibrateseismicresponseofgas chargedAsandsandseismicresponseofano sandzone Seismic WellC Seismic WellD Note Modeltracesarenot TrueAmplitude Data Petro AcousticModeling Petro acousticModelingofReservoirCharacteristicsPerturbingSeismicResponsefromWellCusingsandthickness reservoirporosity andfluidcontentasvariableparameters DecreasedSandThickness DecreasedPorosity DecreasedGasSaturation Petro AcousticModeling PerturbationofReservoirCharacteristicsThicknessvariationsaremodeledfromCtoDwells 70 to0 Synthetictracesandseismictracesaresimilar DecreasedThickness 70 to0 Flattenedseismicsection Intervalofinterest W E Petro AcousticModeling PerturbationofReservoirCharacteristicsSyntheticmodeltracesaregeneratedbetweenCandDwellsusingacombinationofreservoirthickness watersaturationandporosity Syntheticsarethenusedtopilottheseismicfaciesanalysis Syntheticmodeltraces SeismicFaciesMap Post MortemAnalysisofWellD RevisedGeologicalModelSyntheticClassifyingthe82 msecintervalbelowthe despiked referencehorizonusing20modeltraces ThisnewunpilotedseismicfaciesmaphighlightstheoutlineofalateshaleplugfanthatcouldexplaintheabsenceofAsandintheDwell SeismicFaciesMap Modeltraces BusinessImpacts CosttoExposePayDrillingprogramisdirectlyrelatedtothegeologicalmodelMaximizeProductionRateIdentifythesweetspotsHigh gradeprospects Conclusions ExplorationwithintheElkhornSloughfieldhadbeenmostlydrivenbyamplitudeanomalies inversiontechniquesandcoherencytechnology SeismicFaciesAnalysiscombinedwithlitho seismicmodelingofwelldatawasappliedtotheElkhornSloughField Thismethodologyisaccurate cost effective quickandoftenrevealssubtlegeologicalfeaturesonlyexpressedintheshapeoftheseismictrace ThegeologicalmodelwastestedwiththeEwell whichfound100 ofgassand Seismicfaciesmap presentwork Coherencyslice K Lanning1998 TurbiditeCharacterizationUsingMulti AttributeVolumeClassification DataOffshoreAngola 3DSurvey OffshoreAngola Africa650in lines 650cross lines 6 25X6 25binspacingSampleinterval2msec Volumeusedfrom2500 2900msGeneratedAttributes Amplitude Dip Azimuth AI PorosityandSemblanceWellData2wellsdrilledonaturbiditeplayWell2goodproducerfrommassivesandsWell4badproducerfrompoorlysortedsands PartTwo VolumeClassification TheData OffshoreAngola PostStack3DdataStructuralSetting ExtensionalFaultingDepositionalSetting TurbiditeSlopeChannelsStratigraphicInterpretationClassicSeismicTraceShapeAnalysisAttributeMapClassificationMulti AttributeVolumeVisualizationMulti AttributeVolumeClassificationSubvolumeDetectionin3DenvironmentConclusions StructuralSetting TopSystem blue TopSequenceA blue TopChannelA violet IntraChannelA yellow BaseChannelA red ErosionBaseSequenceA green WellswithGRLog Well2isagoodproducerinbothupperandlowerAUnit Well4isapoorproducerfromupperAunitandshowsalowGRresponseforthesecondunitofcorrespondingmassivesandsinWell2 Inline2258Well4 Inline2120Well2 StratigraphicInterp